Much current interest exists in treating cancer with agents that may reduce tumor blood flow through reducing angiogenesis. In addition, tumor blood flow heterogeneity is known to be related to poor treatment response in malignant tumors. Despite the importance of tumor blood flow, there has been little systematic study of the reproducibility of minimally invasive tumor blood flow measurement studies using positron emission tomography (PET) in untreated cancer patients. The purpose of this investigation is to evaluate the precision of repeated measurements of tumor blood flow with (oxygen -15)-water (0-15-water) in untreated human lung cancers.